Affective States in Behavior Networks
2009 (English)In: Intelligent Computer Graphics / [ed] Dimitri Plemenos, Springer , 2009, 1, Vol. 240, 19-39 p.Chapter in book (Other academic)
Several studies within the field of neuroscience, psychology and economics have concluded that emotions are vital for the human decision making process. Without emotions, humans cannot make reasonable choices, a fact that has until recently been neglected within the area of action selection. We propose a method to extend behavior networks to include affective impact. The work by Loewenstein et al. will be our main motivation. In our model the parameters of the behavior network change dynamically according to the agents emotional state and the goals of the network are entirely emotional. The agents risk judgement also changes depending on its current emotions. The aim is to create a platform for modeling human-like affective decision making. Our goal is to create natural and interesting non-player characters for games or game-like applications.
Place, publisher, year, edition, pages
Springer , 2009, 1. Vol. 240, 19-39 p.
, Studies in Computational Intelligence, ISSN 1860-949X ; Vol. 240
virtual agents, decision making, emotions, behavior networks
Engineering and Technology
IdentifiersURN: urn:nbn:se:liu:diva-53444DOI: 10.1007/978-3-642-03452-7_2ISI: 000273182900002ISBN: 978-3-642-03451-0 (print)ISBN: 978-3-642-03452-7 (online)OAI: oai:DiVA.org:liu-53444DiVA: diva2:288843